Posts

CarCrash1

Automated Driving Tech Could End Rise in Road Deaths

Burney Simpson

Deaths in traffic accidents are on the rise but these fatalities could be reduced if automated technology were installed in passenger vehicles.

That’s the conclusion after reading the major findings in two recent reports –

  • Road deaths rose about 8 percent in 2015;
  • Adding three types of currently-available automated technology to passenger cars could reduce accidents by about 25 percent.

Starting with bullet two, a study from Carnegie Mellon University (CMU) found that the installation of blind spot monitoring, lane departure warning, and forward collision warning systems could stop or make less severe 1.3 million crashes a year.

Those crashes cause 10,000 fatalities and 133,000 injuries annually, the researchers report.

It would cost about $600 per light-duty vehicle to install the technology, according to “Cost and benefit estimates of partially-automated vehicle collision avoidance technologies” by Corey D. Harper, Chris T. Hendrickson, and Constantine Samaras. The three are with CMU’s Civil and Environmental Engineering department.

It would cost about $13 billion to install this technology in all light-duty vehicles in the U.S. but this investment would bring an $18 billion benefit in the first year alone.

TRAFFIC FATALITIES RISING

The CMU study comes just as the National Highway Traffic Safety Administration (NHTSA) reports that 2015 traffic fatalities rose nearly 8 percent from 2014 following decades of decline.

An estimated 35,200 people died in traffic accidents in 2015, while total vehicle miles traveled (VMT) rose 3.5 percent to 107.2 billion miles, according to NHTSA.

The fatality rate per 100 million VMT last year rose to 1.12, up from 1.08 in 2014.

Those who weren’t even in passenger vehicles saw “significant increases” in deaths from traffic accidents, NHTSA found.

For instance, traffic fatalities rose 13 percent for bicyclists, 10 percent for pedestrians and 9 percent for motorcyclists last year, while fatalities of drivers and passengers rose by 6 percent and 7 percent respectively.

The 2015 findings remain estimates. An annual statistical report will be released later this year.

The numbers for last year run counter to long-term trends. From 1973 to 2013 crash fatalities dropped about 40 percent due to the use of seat belts, the installation of air bags, education campaigns on the dangers of drunk driving, and greater police enforcement.

However, these improvements have largely stayed the same since 2009 (See “Road Safety Hits a Plateau: Fed Traffic Stats“).

Photo: Flipped car at 22nd and Hawthorne by Aaron Parecki, 2010.

Self-Driving, Electric Forklifts Automate Warehouses

Automated, electric-powered forklifts developed at the Carnegie Mellon Robotics Institute are contributing to the growing use of self-driving technology.

Coraopolis, Penn.-based Seegrid takes off-the-shelf electric-powered forklifts, adds hardware and software, and turns them into Vision Guided Vehicles (VGV), said Amanda Merrell, Seegrid marketing director.

This video shows the VGVs in action and contrasts their measured approach to the driving style of their human counterparts.

Once transformed, the VGVs perform such warehouse tasks as Put Away; Long Haul; End of Line where the finished product is taken to a shipping dock; Replenishment where inventory is moved from storage to picking; and more.

Seegrid says its VGVs reduce labor costs, improve warehouse safety, and increase productivity.

Customers include Volvo, Daimler, Denso, BMW, Jaguar Land Rover, and Freightliner.

The hardware that Seegrid adds includes 10 cameras that keep the forklift aware of its surroundings. The software includes coding that allows the warehouse operator to program a driving route for the VGV.

“You simply get on the machine, hit record, drive it on some particular route, get off, hit play, and it just loops around on that route forever,” Seegrid CEO Jim Rock told WESA.fm in Pittsburgh.

It hasn’t all been easy. Seegrid declared Chapter 11 bankruptcy in 2014 as it got ahead of itself and explored new markets. Rock helped the firm to re-focus on its core forklift sector. In addition, supermarket operator Giant Eagle, an original Seegrid funder, agreed to swap debt for a greater equity share of the firm.

The company exited bankruptcy in February 2015.

A customer that purchases VGVs will use the Seegrid Supervisor app to manage the devices. It provides for real-time monitoring, intersection control, and the ability to set operating rules, such as recharging.

The operator programs the forklift to visit a charging station once its battery power has dropped to a certain level, “say when it’s at 30 percent,” said Merrell.

A human is still needed to physically attach the forklift to the recharging station, said Merrell.

Rock says the VGVs have operated safely for more than 125,000 miles due to sensors that stop the forklift when it senses a human is within a certain distance. And the VGVs beep and make other noises as they move so workers are made aware of their presence.  

Dr. Hans Moravec developed Seegrid’s technology at Carnegie Mellon in Pittsburgh. It was officially founded in 2003 and launched its first VGV in 2008.

Seegrid2

Self-Driving Forklifts Automate Warehouses

Burney Simpson

Automated, electric-powered forklifts developed at the Carnegie Mellon Robotics Institute are contributing to the growing use of self-driving technology.

Coraopolis, Penn.-based Seegrid takes off-the-shelf electric-powered forklifts, adds hardware and software, and turns them into Vision Guided Vehicles (VGV), said Amanda Merrell, Seegrid marketing director.

Once transformed, the VGVs perform such warehouse tasks as Put Away; Long Haul; End of Line where the finished product is taken to a shipping dock; Replenishment where inventory is moved from storage to picking; and more.

Seegrid says its VGVs reduce labor costs, improve warehouse safety, and increase productivity.

Customers include Volvo, Daimler, Denso, BMW, Jaguar Land Rover, and Freightliner.

This YouTube video cleverly contrasts the measured style of the automated VGV with its human-operated counterparts - Seegrid Vision Guided Vehicles in Action.

The hardware that Seegrid adds includes 10 cameras that keep the forklift aware of its surroundings.

The software includes coding that allows the warehouse operator to program a driving route for the VGV.

“You simply get on the machine, hit record, drive it on some particular route, get off, hit play, and it just loops around on that route forever,” Seegrid CEO Jim Rock told radio station WESA fm in Pittsburgh.

It hasn’t all been easy. Seegrid declared Chapter 11 bankruptcy in 2014 as it got ahead of itself and explored new markets. Rock helped the firm to re-focus on its core forklift sector. In addition, supermarket operator Giant Eagle, an original Seegrid funder, agreed to swap debt for a greater equity share of the firm.

The company exited bankruptcy in February 2015.

A customer that purchases VGVs will use the Seegrid Supervisor app to manage the devices. It provides for real-time monitoring, intersection control, and the ability to set operating rules, such as recharging.

To do that, the VGV is programed to visit a charging station once its battery power has dropped to a certain level, “say when it’s at 30 percent,” said Merrell.

A human is still needed to physically attach the forklift to the recharging station, said Merrell.

Rock says the VGVs have operated safely for more than 125,000 miles due to sensors that ‘tell’ the machine when a human is within a certain distance. And the VGVs beep and make other noises as they move so workers are made aware of their presence.  

Dr. Hans Moravec developed the Seegrid technology at Carnegie Mellon in Pittsburgh. The company was founded in 2003 and launched its first VGV in 2008.

S-701_med

Autonomous Helicopter Deploys Autonomous Ground Vehicle: Video

This video shows two autonomous vehicles collaborating as the U.S. military tests an autonomous helicopter and an autonomous Land Tamer II ground vehicle.

This test was conducted by Carnegie Mellon’s National Robotics Engineering Center and Sikorsky Aircraft. The ground vehicle uses Velodyne LiDAR.

The technology could be used for tactical or long-distance resupply, risk assessment, reconnaissance and similar missions.

GrrenLites1

Green Lights All the Way

Burney Simpson

It’s green lights all the way baby in parts of Pittsburgh due to the implementation of a decentralized traffic system overseen by a Carnegie Mellon robotics professor.

The adaptive traffic control signal system from Surtrac uses a combination of traffic theory and artificial intelligence to follow traffic flow on a real-time basis.

The signals at intersections are controlled by a camera and computer system that watches the traffic coming from every direction, said Stephen F. Smith, research professor and director of the Intelligent Coordination and Logistics Laboratory in the Robotics Institute at Carnegie Mellon.

Surtrac can change the lights to green when there is no cross-traffic coming, and allows nearby intersections to communicate with each other so drivers can hit a series of green lights when the timing is right.

“It sets a phase system for green lights, and sends the information to other intersections,” said Smith, who recently presented his findings at the 9th University Transportation Center Spotlight Conference in Washington, D.C. organized by the Transportation Research Board.

Surtrac’s decentralized signal control contrasts with traffic oversight in most urban areas where signal timing is determined by central planners using studies of past traffic flow.

The system has been installed at 50 intersections in Pittsburgh’s East End neighborhood. Since it began in 2012, the project has cut travel time by 24 percent and waiting time by 42 percent, according to office of the Pittsburgh mayor.

That efficiency has led to a 21 percent reduction in emissions, said Smith who leads Rapid Flow Technologies, a Carnegie Mellon spin off. The university owns a piece of Rapid Flow as it seeks to commercialize Surtrac.

Rapid Flow is also researching ways to detect and measure the frequency of cruising for parking spaces, and building a system that measures the performance of surfaced road networks.

Smith would like to expand the Surtrac program to integrate vehicles, bikes, and pedestrians to better shape the flow of all traffic.

Pittsburgh has asked Rapid Flow to research whether Surtrac would work in the busy North Shore area, home to major sports stadiums Heinz Field and PNC Park, along with the Andy Warhol Museum.

Photo: Green traffic light when it’s raining by Sigfrid Lundberg, 2014.